
Re: Concrete nonlinear elastic
bill campbell Feb 19, 2014 12:27 PM (in response to Mousa D)Sounds strange. Here's a couple of things to try:
1. In the cosmos (simulation) folder, RMB and click "Update All Components". This will for resync the cosmos and SW components. In the old days, we saved and reopened the part to update components. May still have to do this.
2. Make sure you make a copy of the material and paste it before changing any of the values. You can not change the library.
Which Model Type are you using? nonlinear elastic?

Re: Concrete nonlinear elastic
Shaun Densberger Feb 19, 2014 12:34 PM (in response to Mousa D)"...have noticed that even If I increased the pressure by x10 I'd still get same displacement and stress..."
Can you post an image of how your model is set up (geometry, boundary conditions, loads, etc.)
"Then again, going back to materials properties, when I change the yield strength from the previous 4.35N/mm2 I have set to 1N/mm2 .. still no changes at all."
Yield strength does not impact the numeric values for a linear elastic analysis. Yield strength is the point at which the material yields; the Young's Modulus defines the "stiffness" of the material. If you modify that value then you'll see different numeric values.

Re: Concrete nonlinear elastic
Jared Conway Feb 19, 2014 1:55 PM (in response to Mousa D)1. only E is going to change the way the model behaves, and v in some cases. post some pictures of the displacements and stresses of both setups and also take a look at the reaction force on your restraint. since this sounds like a simple problem and user error to start, work on a simple block and post the model with both studies setup.
2. what is the exact workflow that causes this problem? once you set a material, it will not change back to linear elastic. just make sure that you're using a nonlinear study.

Re: Concrete nonlinear elastic
Mousa D Feb 20, 2014 5:39 AM (in response to Mousa D)Thank you, really appreciate the quick replies. My questions are as follows:
 Using my first study, which is linear elastic isotropic for a concrete material that I have defined. (Included a screen shot below) As I am interested in the Yaxis Stresses, the results I have obtained seemed fairly reasonable. Yet, how can I be given results and different values for the Von Mises, even though it is set as linear elastic, surely there will not be plastic (Von mises) in my case? As a Von mises, I get yield strength as 4.3MPa, would that be my yield strength in the Yaxis stresses as well?
 In my second study, I have changed the simulation into nonlinear, and then went to the concrete material properties and changed it into nonlinear elastic, and created stressstrain that is fairly close enough to a concrete stressstrain curve. Only the compression part, or would I be able to draw the stressstrain for the tension part as well?
 What is the difference between editting my material property to nonlinear elastic, and running my study as nonlinear elastic. As when I press apply material, it shows as nonlinear elastic, yet when I right click on edit material on the left column of the simulation, it shows it as linear elastic still. I have defined my stress strain curve as following, roughly estimated it, if there is a file or values that is more accurate to put in, please let me know.
 I have not ran the nonlinear simulation yet as it apparently would take more than 5 hours, as I have tried it and it took about 30 minutes to reach 6%, so I cancelled it and wanted to be sure of all properties and such before running my simulation.
(My yield strength/tensile strength is 4.35N/mm2)

Re: Concrete nonlinear elastic
Mousa D Feb 20, 2014 5:42 AM (in response to Mousa D)I have copy/pasted 4 picture in my above post, I am not sure if it is visible to you or not.

Re: Concrete nonlinear elastic
Martijn Loonen Feb 20, 2014 6:05 AM (in response to Mousa D)Not visible, you have to upload the images.
I like the new snipping tool in windows 7 to take partial screenshots. (under accessory programs)


Re: Concrete nonlinear elastic
Jared Conway Feb 20, 2014 11:35 AM (in response to Mousa D)Using my first study, which is linear elastic isotropic for a concrete material that I have defined. (Included a screen shot below) As I am interested in the Yaxis Stresses, the results I have obtained seemed fairly reasonable. Yet, how can I be given results and different values for the Von Mises, even though it is set as linear elastic, surely there will not be plastic (Von mises) in my case? As a Von mises, I get yield strength as 4.3MPa, would that be my yield strength in the Yaxis stresses as well?
>> you're expecting y stress to be the same as von mises?
 In my second study, I have changed the simulation into nonlinear, and then went to the concrete material properties and changed it into nonlinear elastic, and created stressstrain that is fairly close enough to a concrete stressstrain curve. Only the compression part, or would I be able to draw the stressstrain for the tension part as well?
>> have you checked the documentation and the solidworks KB? i believe this is possible but it would be covered in these locations.
 What is the difference between editting my material property to nonlinear elastic, and running my study as nonlinear elastic. As when I press apply material, it shows as nonlinear elastic, yet when I right click on edit material on the left column of the simulation, it shows it as linear elastic still. I have defined my stress strain curve as following, roughly estimated it, if there is a file or values that is more accurate to put in, please let me know.
>>what left column?
>>file or values that are more accurate? this would come from your knowledge of the application. only you can answer this question.
 I have not ran the nonlinear simulation yet as it apparently would take more than 5 hours, as I have tried it and it took about 30 minutes to reach 6%, so I cancelled it and wanted to be sure of all properties and such before running my simulation.
>>this sounds like a good idea. my recommendation would be to start with just a block and get a hang of how solidworks simulation works first. go through the tutorials, go through the help and thne if you can, take your reseller's training to get a good introduction to the interface. and if they have it, go through some oneonone training. here at hawk ridge systems we offer both to our customers to give them the best chance of quick success with solidworks simulation.

Re: Concrete nonlinear elastic
Shaun Densberger Feb 20, 2014 2:15 PM (in response to Mousa D)"As I am interested in the Yaxis Stresses, the results I have obtained seemed fairly reasonable. Yet, how can I be given results and different values for the Von Mises, even though it is set as linear elastic, surely there will not be plastic (Von mises) in my case? As a Von mises, I get yield strength as 4.3MPa, would that be my yield strength in the Yaxis stresses as well?"
There's many different failure theories used in solid mechanics, but the two most common used in general FEA are:
 VonMises Yield Criteria
 Mohr–Coulomb Theory
One key difference between the two is that vonMises is good for ductile materials (steel, aluminum, etc.), while MohrCoulomb is good for brittle materials (cast iron, concrete, etc). Another important note is that vonMises is a scalar quantity, meaning that it doesn't have a direction associated with it (hence a negative vonMises stress being meaning).
"In my second study, I have changed the simulation into nonlinear, and then went to the concrete material properties and changed it into nonlinear elastic, and created stressstrain that is fairly close enough to a concrete stressstrain curve. Only the compression part, or would I be able to draw the stressstrain for the tension part as well?"
I've never really worked with brittle materials before, but I know that for ductile materials (like steel), the strainstrain curve is the same for tension or compression within the elastic region, but different in the plastic region. This link is worth reading. http://www2.ce.metu.edu.tr/~ce581/docs/Lecture%20Notes/Chapter%201/1_1_MATERIAL_BEHAVIOR.pdf
"What is the difference between editting my material property to nonlinear elastic, and running my study as nonlinear elastic. As when I press apply material, it shows as nonlinear elastic, yet when I right click on edit material on the left column of the simulation, it shows it as linear elastic still. I have defined my stress strain curve as following, roughly estimated it, if there is a file or values that is more accurate to put in, please let me know."
I don't use SW simulation that much, but I'm almost positive that if you want to run a nonlinear material model, you need to run a nonlinear analysis, which mean creating a new analysis and selecting nonlinear static. Others can correctly me if I'm wrong.
"I have not ran the nonlinear simulation yet as it apparently would take more than 5 hours, as I have tried it and it took about 30 minutes to reach 6%, so I cancelled it and wanted to be sure of all properties and such before running my simulation."
Just be warned that nonlinear can be rather difficult and somewhat trial and error based when you're starting out with it.

Re: Concrete nonlinear elastic
Mousa D Feb 20, 2014 8:40 AM (in response to Mousa D)Here is a file that contains all pictures taken, Linear, nonlinear, and the error I get when I run nonlinear. (As a PDF)

Error, nonlinear elastic.pdf 1,020.2 K


Re: Concrete nonlinear elastic
Shaun Densberger Feb 20, 2014 2:21 PM (in response to Mousa D)I looks like you've defined a nonlinear dynamic analysis. While you could solve this type of problem with nonlinear dynamics, I would recommend that you stick with nonlinear static. Try you're model again was a nonlinear static analysi. If the analysis fails, try cutting your load (try 1% of your current load) and see if it'll solve. If it doesn't, then try replacing your pressure load with an enforced displacement and rerunning the analysis. Let us know the results of all of this.

Re: Concrete nonlinear elastic
Jared Conway Feb 20, 2014 4:14 PM (in response to Shaun Densberger)I mousa, i'd concentrate on starting simple.
get the hang of a linear static, understand what it means and the assumptions that go along with it.
create posts for each issue you run into and try attacking them one at a time (many to one vs one to many is a lot easiest to approach)
then build up as your knowledge increases
also if you can, i'd highly recommend involving your reseller who can provide training in the basics of simulation and then technical support in a quicker fashion than through the forum

Re: Concrete nonlinear elastic
Mousa D Feb 21, 2014 5:57 AM (in response to Shaun Densberger)Hello Jared, thank you for your reply. I have defined a nonlinear static analysis, and then ticked the 2D simplification but did not understand what values I should put in, so unticked it and continued. As you have mentioned it failed when I've put my pressure as 2.7N/mm2 from the first iteration saying "PCGLSS0139:Matrix is Singular or Indefinite*****************" , then changed my pressure into 2.7N/m2 and it solved really quick. (Even though the stresses in Yaxis that I am interested in were all zeros.) What would be my problem and how can I fix it please?
Many thanks in advance.

Re: Concrete nonlinear elastic
Jared Conway Feb 21, 2014 12:08 PM (in response to Mousa D)matrix is singular = not enough restraints in your problem or your load is too high potentially
going to nonlinear and 2d without having the fundamentals of static is probably relatively challenging. i'd recommend starting with a block in a static analysis. get that to work. then you can move from the block to your model with a fixed restraint. then your loading. then nonlinear. if you want to throw 2d in the mix. go from block to 2d and then your geometry..etc.



Re: Concrete nonlinear elastic
Jared Conway Feb 20, 2014 4:14 PM (in response to Mousa D)what are you trying to show in this doc?


Re: Concrete nonlinear elastic
Mousa D Feb 24, 2014 5:12 PM (in response to Mousa D)Hello, I managed to over come few of the problems I had earlier. Defining my own material (Concrete) with reference to steel, I was interested in the tension curve, knowing that for steel the tension and compression stressstrain curves are equal, I defined compression stressstrain curve as the tension one. Just for reference, does the software enable you to make both the stressstrain curve for tension and compression, or is it just compression?

Re: Concrete nonlinear elastic
Jared Conway Feb 24, 2014 5:38 PM (in response to Mousa D)what material model are you using? most of them are well documented in the help and the kb

Re: Concrete nonlinear elastic
Mousa D Feb 24, 2014 6:21 PM (in response to Jared Conway)I have used the Cast Carbon Steel

Re: Concrete nonlinear elastic
Jared Conway Feb 24, 2014 6:26 PM (in response to Mousa D)is it a linear material? nonlinear elastic? plasticity von mises?

Re: Concrete nonlinear elastic
Mousa D Feb 24, 2014 6:44 PM (in response to Jared Conway)For the study I am running right now, It is a nonlinear elastic study, my stressstrain curve is a bilinear curve. Definied my own material properties. (Which is concrete, 0.2 poisson's ratio, 2300 Kg/m3 density, 3.5N/mm2 tensile strength)

Re: Concrete nonlinear elastic
Jared Conway Feb 24, 2014 11:57 PM (in response to Mousa D)http://help.solidworks.com/2014/English/SolidWorks/cworks/c_Nonlinear_Elastic_Model.htm
have you taken a look at the help?
or the solidworks kb?
S021165: Can I have different material behavior in compression and tension? Only elastic MPC curves can be defined to have a different behavior in compression and tension. Plastic MPC curves do not support a definition in the compression zone. Source: Solutions 
Re: Concrete nonlinear elastic
Mousa D Feb 26, 2014 5:47 PM (in response to Jared Conway)Had a look at the help option, but unfortunately I can not access the kb, or the link you have provided at the end, as I am a student in a University and we do not have our own registration number or so.
Following to my model, I am adding a simple UDL pressure accross the front third of the rectangle I am facing. Getting the whole structure in red, with a really small displacement, I am getting an odd 1 mm displacement on one point of my structure, and I can not seem to understand why?

Solidworksodd.pdf 253.6 K

Re: Concrete nonlinear elastic
Jared Conway Feb 26, 2014 7:45 PM (in response to Mousa D)looks like a setup problem. does it happen on a simple block?

Re: Concrete nonlinear elastic
Jared Conway Feb 26, 2014 7:46 PM (in response to Jared Conway)could also be mesh i guess. are you using draft quality elements? is your mesh converged?

Re: Concrete nonlinear elastic
Jared Conway Feb 26, 2014 7:46 PM (in response to Jared Conway)how does it behave with linear materials? what about a linear static?

Re: Concrete nonlinear elastic
Mousa D Feb 27, 2014 5:07 AM (in response to Jared Conway)Hello Jared, yes I am using a standard mesh, draft quality mesh, and set it as the coarsest.

Re: Concrete nonlinear elastic
Jared Conway Feb 27, 2014 3:15 PM (in response to Mousa D)so where are you stuck still?

Re: Concrete nonlinear elastic
Mousa D Mar 2, 2014 3:33 PM (in response to Jared Conway)Well, I was wondering, since Im doing some testing on a certain model, I am using exaclty the same parameters for each test, yet Im trying to see the difference between each one of them. In each test all I have changed is the meshing element, starting from Finest to Coarsest. I have attached a file to my readings.
I was confused about the time stepping. As you would be able to see from my table, the max time step reached, out of 5 seconds for the finest is around 3.8 as it starts increasing till it reaches the coarsest where it goes back to 3.7, Shouldnt it be constant? I could not come with an answer of why that is happening. Any advice or explaning would be greatly appreciated.

meshing table.docx 26.0 K

Re: Concrete nonlinear elastic
Jared Conway Mar 2, 2014 10:20 PM (in response to Mousa D)What relevance do you Think it has?
What effect did it have on your results or do you Thibk it has on your results?
The time step is not guaranteed between studies even if they are setup the same, the auto meager is in place to get the most accuse at solution time in the quickest amount of time. So between element sizes and therefore stiffness matrix and solution, it may change.
I also would not recommend using the manual mesh size unless you've been told to do that.

Re: Concrete nonlinear elastic
Mousa D Mar 3, 2014 3:39 AM (in response to Jared Conway)Good morning Jared,
Thank you again for your reply. I understand that It all affects and comes down to the original Kd=F equation. Can you please explain more. As you mentioned earlier, the default of the mesh of my structure will be automatically set to reach the highest time step before failing? Before obtaining my results, I would have thought the finer I make my mesh the more solution time my steps would reach.
That is to say, for example, If I had set it on coarse or higher size than the default, my last solution time out of 5 seconds for example would be 4 out of 5. Yet, if decreased the element size as to be finer, wouldnt it reach a higher solution time? 4.3 out of 5 for example.
Thank you again in advance and for all explanations I will kindly receive.
Regards,
M.Dahabreh.

Re: Concrete nonlinear elastic
Jared Conway Mar 3, 2014 9:58 AM (in response to Mousa D)If your problem is setup correctly and is numerically solvable, your solution will completely regardless of mesh size.
If what you are calling max step size is the maximum length of the solution, then we aren't on the same page at all. I would recommend you look through the tutorials, help and take the nonlinear class from your reseller. We may have an online one coming up if you don't have one from your reseller soon.
That being said, if your problem is still failing to complete, you should be looking at the setup to understand why it is not solvable. This may be best done by your reseller as it may need attention by the developers. But feel free to post here and someone can take a look. But note, previously your problem still had not addressed some of the main suggestions like starting simple with linear static..etc which is highly recommended when you are starting.













